GitHub
nnsight-basics
Core nnsight concepts for neural network interpretability. Use when setting up models, tracing activations, saving values, or making basic interventions on model internals.
Skills IA tools llm ai
Découvrez les skills tools llm ai, comparez leurs dépôts et consultez leurs instructions.
Catégories: tools llm ai
tools llm ai Effacer les filtres
Explorer
GitHub
perplexity
AI-powered web search and reasoning using Perplexity API through the scraper MCP server. Invoked when searching the web with AI synthesis, conducting research with citations, performing complex reasoning tasks,...
GitHub
socratic-debate
This skill should be used when conducting structured multi-perspective debates to stress-test ideas, evaluate tradeoffs, or reach well-reasoned decisions. Provides the Socratic debate framework with 4 subagent...
GitHub
check-skill-conflicts
This skill should be used when checking for naming conflicts between local skills (~/.claude/skills) and plugin-provided skills (~/.claude/plugins). Use to identify duplicate or similarly named skills that may...
GitHub
reagent-review
Use this when you want to create or get a Reagent Review Session. Example requests: 'Lets create a review session for the changes youve made' 'Create a review session for the plan you wrote in the md file' 'Get...
GitHub
scrum-team-developer
AI Developer following TDD principles in AI-Agentic Scrum. Use when implementing PBIs, managing subtasks, or executing the TDD cycle within Scrum.
GitHub
claudecraft
Spec-driven development workflow skill. Auto-loads when working on: - Specification creation or editing - Technical planning - Task decomposition - Implementation within ClaudeCraft context
GitHub
claudecraft
Spec-driven development workflow skill. Auto-loads when working on: - Specification creation or editing - Technical planning - Task decomposition - Implementation within ClaudeCraft context
GitHub
prompt-engineering
Use when writing or improving any LLM prompts. Applies TDD methodology and research-backed practices (Meincke 2025): test don't assume, measure baseline, iterate rigorously. Prevents assuming universal techniqu...
GitHub
skill-writing
Required when creating SKILL.md files for Claude Code agents. Specializes prompt-engineering TDD with agent-specific requirements: frontmatter format, persuasion principles, 500-line limit, progressive disclosu...
GitHub
claude-code-optimizer
Claude Code productivity patterns including EPCC workflow, context management, and extended thinking triggers. Use when starting sessions or managing context.
GitHub
skill-creator
Use when creating a new skill, updating an existing skill, or learning skill best practices. Load for extending Claude's capabilities with specialized workflows, tool integrations, or domain expertise. Covers s...
GitHub
browser-testing
Use when testing web applications, debugging browser console errors, automating form interactions, or verifying UI implementations. Load for localhost testing, authenticated app testing (Gmail, Notion), or reco...
GitHub
claude-md-creator
Use when creating CLAUDE.md files, updating existing CLAUDE.md, validating CLAUDE.md structure, or auto-fixing CLAUDE.md issues. Load for setting up project instructions, global guidelines, local overrides, or...
GitHub
initialization
Use when starting a new session without feature-list.json, setting up project structure, or breaking down requirements into atomic features. Load in INIT state. Detects project type (Python/Node/Django/FastAPI)...
GitHub
context-graph
Use when storing decision traces, querying past precedents, or implementing learning loops. Load in COMPLETE state or when needing to learn from history. Covers semantic search with Voyage AI embeddings, Chroma...
GitHub
determinism
Use when verifying outcomes with code instead of LLM judgment, versioning prompts with hashes, or ensuring reproducible agent behavior. Load for any critical verification. Scripts return boolean exit codes, not...
GitHub
token-efficient
Use when processing 50+ items, analyzing CSV/log files, executing code in sandbox, or searching for tools. Load for data processing tasks. Achieves 98%+ token savings via in-sandbox execution, progressive discl...
GitHub
two-agent-harness
This skill sets up a complete two-agent development system based on the "Effective Harnesses for Long-Running Agents" research. It creates initializer-agent (for project planning) and coding-agent (for incremen...
GitHub
mcp-builder
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or...
GitHub
model-selection
Select appropriate Claude model (Opus 4.5, Sonnet, Haiku) for agents, commands, or Task tool invocations based on task complexity, reasoning depth, and cost/speed requirements.
GitHub
prompt-engineering
Apply prompting techniques when creating prompts, agents, commands, system instructions, or SKILL.md files. Use for XML tags, multishot examples, chain-of-thought, response prefilling, and Claude 4-specific pat...
GitHub
knowledge-synthesis
Synthesize knowledge from multiple sources into Zettelkasten notes for Logseq. Use when creating wiki pages, integrating academic research, or building interconnected knowledge with [[links]] and
GitHub
research-workflow
Apply systematic research methodology for multi-step research, fact-finding, web search, or verification tasks. Use when performing Brave Search, Puppeteer navigation, or synthesizing information from multiple...
Le classement utilise les étoiles du dépôt GitHub indexé. Plusieurs skills peuvent partager ce nombre : ce n’est ni un nombre d’installations ni une note de qualité.